I feel like at some level, it will be much easier to just pir.... I mean... train LLMs based on their content. Yeah. LLM training. That's acceptable. So it really doesn't matter who wins, we'll just perform LLM training.